CXCL12 Neutralizing Antibody Promotes Hair Growth in Androgenic Alopecia and Alopecia Areata

Abstract

We had previously investigated the expression and functional role of C-X-C Motif Chemokine Ligand 12 (CXCL12) during the hair cycle progression. CXCL12 was highly expressed in stromal cells such as dermal fibroblasts (DFs) and inhibition of CXCL12 increased hair growth. Therefore, we further investigated whether a CXCL12 neutralizing antibody (αCXCL12) is effective for androgenic alopecia (AGA) and alopecia areata (AA) and studied the underlying molecular mechanism for treating these diseases. In the AGA model, CXCL12 is highly expressed in DFs. Subcutaneous (s.c.) injection of αCXCL12 significantly induced hair growth in AGA mice, and treatment with αCXCL12 attenuated the androgen-induced hair damage in hair organ culture. Androgens increased the secretion of CXCL12 from DFs through the androgen receptor (AR). Secreted CXCL12 from DFs increased the expression of the AR and C-X-C Motif Chemokine Receptor 4 (CXCR4) in dermal papilla cells (DPCs), which induced hair loss in AGA. Likewise, CXCL12 expression is increased in AA mice, while s.c. injection of αCXCL12 significantly inhibited hair loss in AA mice and reduced the number of CD8⁺, MHC-I⁺, and MHC-II⁺ cells in the skin. In addition, injection of αCXCL12 also prevented the onset of AA and reduced the number of CD8⁺ cells. Interferon-γ (IFNγ) treatment increased the secretion of CXCL12 from DFs through the signal transducer and activator of transcription 3 (STAT3) pathway, and αCXCL12 treatment protected the hair follicle from IFNγ in hair organ culture. Collectively, these results indicate that CXCL12 is involved in the progression of AGA and AA and antibody therapy for CXCL12 is promising for hair loss treatment.

Keywords: CXCL12; alopecia areata; androgenic alopecia; dermal fibroblasts; neutralizing antibody.

Figure 1

Neutralization of CXCL12 promotes hair growth in a testosterone-induced AGA model. (A) The back skin of 7-week-old C₃H male mice was shaved and subcutaneous (s.c.) injections of testosterone propionate (TP) were used to induce the AGA. (B) The expression of CXCL12 in the dorsal skin of control and TP-treated mice was detected using immunofluorescence. CXCL12⁺ cells (red) are indicated by white arrows, and DAPI staining (blue) indicates cell nuclei. The scale bar is set at 100 μm. (C) Different doses of CXCL12 neutralizing antibody (αCXCL12, 5, and 20 μg) were subcutaneously injected twice a week for three weeks. A daily topical treatment of 0.05% dutasteride served as a positive control. αCXCL12 administration accelerated hair growth in TP-treated C₃H mice and the hair weight was measured. ### p < 0.001 vs. control, * p < 0.05, ** p < 0.01 vs. TP-treated, n = 7. (D) αCXCL12 (10, 100, or 1000 ng/mL) treatment increased the length of mouse vibrissa follicles in the AGA mimic ex vivo model. ## p < 0.01 vs. control, * p < 0.05 vs. TP-treated, n = 8. + or -: treated with or without TP, αCXCL12. The asterisk and sharp symbols indicate statistical differences using Student’s t-test.

Figure 2

Androgen treatment enhances CXCL12 expression in dermal fibroblasts. (A) Dermal fibroblasts (DFs) were treated with various concentrations of TP (1, 10, and 100 nM) and DHT (1, 10, and 100 nM) for 24 h, and the expression of CXCL12 was assessed using qRT-PCR. $ p < 0.05 vs. control. (B) After treating DFs with different concentrations of TP and DHT for 48 h, the culture medium was collected and the secreted CXCL12 levels were quantified using ELISA. $ p < 0.05, $$ p < 0.01 vs. control. The dollar sign ($) indicates differences in one-way ANOVA. (C) Immunostaining revealed that the translocation of the AR (green) in DFs increased after TP (100 nM) and DHT (100 nM) treatment for 1 h, as indicated by the white arrows. DAPI staining (blue) marks the cell nuclei. The scale bar is set at 50 μm. (D) After AR-CRISPR/Cas9 knockout (AR-KO) for 48 h, DFs were treated with TP and DHT for an additional 48 h to collect the culture medium for ELISA analysis. AR-KO significantly reduced the secretion of CXCL12 from DFs. Western blot analysis indicated differences in AR expression between the control and AR-KO groups. ### p < 0.001 vs. control, ** p < 0.01, *** p < 0.001 vs. TP or DHT treatment. The asterisk and sharp symbols indicate statistical differences using Student’s t-test.

Figure 3

CXCL12 secreted from DFs induces AR and CXCR4 in DPCs. The effects of rCXCL12 on the expression of AR and CXCR4 in DPCs were observed using qRT-PCR (A,B) and Western blot analysis (C). rCXCL12 increased the mRNA and protein expression of the AR and CXCR4 in DPCs. $ p < 0.05, $$ p < 0.01 vs. Control. (D) DFs were treated with 100 nM TP or DHT for 48 h and the culture medium (CM) was collected. CM from DFs treated with TP and DHT (DFCM^TP and DFCM^DHT) significantly reduced hair length in human hair organ culture. $ p < 0.05 vs. DFCM, n = 10. The dollar sign ($) indicates differences in a one-way ANOVA.

Figure 4

CXCL12 neutralization improves alopecia areata. (A) In this experimental design, skin-draining lymph node (SDLN) cells were isolated from AA-affected C₃H/HeN female mice and intradermally injected into the dorsal skin of naïve mice to induce AA. Severe hair loss was observed after 12 weeks, and αCXCL12 (20 μg) was subcutaneously injected twice in a 2-week interval. (B) Skin sections from normal and AA mice were stained with an anti-CXCL12 antibody (green). The expression of CXCL12 increased in the AA model. The scale bar is set at 500 μm. (C) The administration of αCXCL12 in AA mice significantly reduced the areas of hair loss, which are marked in the images with dotted lines. The extent of hair loss in both the control and αCXCL12-treated mice was quantified using Image J (v1.53t). *** p < 0.001 vs. control, n = 3. (D) Skin sections of AA and αCXCL12-treated mice were stained with anti-CD8, anti-MHC-I, or anti-MHC-II antibodies. The αCXCL12-treated group had reduced expression of these immune reaction markers (green; white arrows). DAPI staining (blue) indicates cell nuclei. The scale bar is set at 100 μm. An asterisk indicates a statistical difference using Student’s t-test.

Figure 5

CXCL12 neutralization prevents alopecia areata onset. (A) In this experimental design, 10-week-old C₃H female mice were treated with αCXCL12 (20 μg) through subcutaneous injections once a week for 14 weeks following AA-affected SDLN cell transplantation. (B) αCXCL12 treatment significantly delayed the onset of AA, which was marked with dotted lines. (C) The incidence of AA onset in SDLN-transplanted C₃H mice. Control (n = 10) or αCXCL12 treated (n = 9). (D) Skin sections were stained with an anti-CD8 antibody, and αCXCL12 reduced the expression of CD8⁺ cells (green; white arrows). DAPI staining (blue) indicates the cell nuclei. The scale bar is set at 100 μm. $$ p < 0.01 vs. control. The dollar sign ($) indicates differences in one-way ANOVA.

Figure 6

Involvement of JAK/STAT signaling in IFNγ-mediated CXCL12 secretion. (A) Dermal fibroblasts (DFs) were treated with IFNγ at various concentrations (1, 10, 30, and 100 ng/mL), and the expression of CXCL12 was significantly increased. $ p < 0.05 vs. control. (B) The secreted CXCL12 level was measured using ELISA, and IFNγ significantly increased CXCL12 secretion. $$$ p < 0.001 vs. control. The dollar sign ($) indicates differences in one-way ANOVA. (C) Immunostaining showed that p-STAT3 (green) in DFs was increased by IFNγ (100 ng/mL), as indicated by the white arrows. DAPI staining (blue) indicates the cell nuclei. The scale bar is set at 100 μm. (D) JAK inhibitor (Baricitinib) and STAT3 inhibitor (Stattic) treatment attenuated the IFNγ-induced CXCL12 secretion from DFs. ## p < 0.01 vs. control, * p < 0.05, ** p < 0.01 vs. IFNγ treated. (E) αCXCL12 treatment increased the length of mouse vibrissae follicles in the AA mimic ex vivo model. # p < 0.05 vs. control * p < 0.05 vs. IFNγ treated, n = 8. Asterisk and sharp indicate statistical differences using Student’s t-test.